Download Kienle, Miriam - Between Nodes and Edges

Artl@s Bulletin
Volume 6
Issue 3 Visualizing Networks: Approaches to Network
Analysis in Art History
Article 1
2017
Between Nodes and Edges: Possibilities and Limits
of Network Analysis in Art History
Miriam Kienle
University of Kentucky, [email protected]
Follow this and additional works at: https://docs.lib.purdue.edu/artlas
Part of the Digital Humanities Commons, Fine Arts Commons, Interdisciplinary Arts and Media
Commons, and the Theory and Criticism Commons
Recommended Citation
Kienle, Miriam. "Between Nodes and Edges: Possibilities and Limits of Network Analysis in Art History." Artl@s Bulletin 6, no. 3
(2017): Article 1.
This document has been made available through Purdue e-Pubs, a service of the Purdue University Libraries. Please contact [email protected] for
additional information.
This is an Open Access journal. This means that it uses a funding model that does not charge readers or their institutions for access. Readers may freely
read, download, copy, distribute, print, search, or link to the full texts of articles. This journal is covered under the CC BY-NC-ND license.
Visualizing Networks:
Approaches to Network
Analysis in Art History
Miriam Kienle, Guest Editor
ARTL@S BULLETIN
Volume 6, Issue 3
Fall 2017
Editorial Statement
Guest Editor
The ARTL@S BULLETIN is a peer-reviewed, transdisciplinary
journal devoted to spatial and transnational questions in the
history of the arts and literature.
The Artl@s Bulletin ’s ambition is twofold: 1. a focus on the
“transnational” as constituted by exchange between the local
and the global or between the national and the international; 2.
an openness to innovation in research methods, particularly the
quantitative possibilities offered by digital mapping and data
visualization.
We publish two to three thematic issues every year. If you would
like to contribute to the journal with an article or propose a
theme for a future issue, please contact the editors Catherine
Dossin ([email protected]) and Béatrice Joyeux-Prunel
([email protected]). We welcome suggestions,
ideas, and submissions from scholars worldwide and at every
stage in their career.
Miriam Kienle
University of Kentucky
Copyright Statement
Editors-in-Chief
Catherine Dossin
Purdue University
Béatrice Joyeux-Prunel
École normale supérieure
PSL Research University Paris
Cover: Detail of a Network
diagram by Yael Rice showing
betweenness centrality of
collaborators (designers and
colorists) on the Jaipur
Razmnama (1582/3–6)
manuscript illustrations.
ARTL@S BULLETIN (ISSN 2264-2668) is published by the École
normale supérieure, 45, rue d’Ulm, 75005 Paris, France and the
Centre national pour la recherche scientifique 16, rue Pierre et
Marie Curie, 75005 Paris, France. The online version of the
ARTL@S Bulletin is hosted by Purdue Scholarly Publishing
Services.
Compilation copyright © 2017 ENS-CNRS. Contents copyright
© 2017 by the respective authors, artists, and other rights
holders. All rights in the ARTL@S BULLETIN and its contents
reserved by the École normale supérieure, the Centre national
pour la recherche scientifique and their respective owners.
Except as permitted by the Copyright Act, including section 107
(the fair use doctrine), or other applicable law, no part of the
contents of the ARTL@S BULLETIN may be reproduced without the
written permission of the author(s) and/or other rights holders.
The opinions expressed in the ARTL@S BULLETIN are those of the
authors and not necessarily of the editors or publishers.
Table of Contents
4
Between Nodes and Edges: Possibilities and Limits of Network Analysis in Art History
Miriam Kienle
23
Continuity and Disruption in European Networks of Print Production, 1550-1750
Matthew D. Lincoln
41
Keeping Our Eyes Open: Visualizing networks and art history
Stephanie Porras
50
Workshop as Network: A Case Study from Mughal South Asia
Yael Rice
66
Network Analysis and Feminist Artists
Michelle Moravec
81
The Computer as Filter Machine: A Clustering Approach to Categorize Artworks Based on
a Social Tagging Network
Stefanie Schneider and Hubertus Kohle
90
Enriching and Cutting: How to Visualize Networks Thanks to Linked Open Data Platforms
Léa Saint-Raymond and Antoine Courtin
108
What You See Is What You Get: The “Artifice of Insight.” A Conversation between R. Luke
DuBois and Anne Collins Goodyear
Anne C. Goodyear
121
Digital Art History “Beyond the Digitized Slide Library”: An Interview with Johanna
Drucker and Miriam Posner
Miriam Kienle
Miriam Kienle
University of Kentucky
Between Nodes and Edges: Possibilities and
Limits of Network Analysis in Art History
I
n the 2016 issue of the International
Journal for Digital Art History on
“Visualizing Big Image Data,” the
featured article by the multidisciplinary art
historian Maximilian Schich opens with a still image
from an animated map that visualizes the
movement of notable historical figures from places
of birth to death (blue to red) [Fig. 1].1 Hovering
above the earth, connections become illuminated.
This macroscopic perspective, as Schich states,
rewards viewers with “breathtaking beauty and
radically new insights that cannot be achieved by
local inquiry.”2 Originally produced as part of a
study that Schich’s research team published in the
high-impact journal Science, the visualization was
utilized to locate and understand the quantitative
patterns of history.3 Applying a “network
framework of cultural history,” they analyzed the
migration of more than 150,000 notable individuals
(drawn from three major US and European
databases)4 to find a number of significant patterns,
including clear evidence that modern Europe has
been characterized by two different cultural
systems: a “winner-takes-all regime,” in which
historical actors clustered around centers such as
Paris, and a “fit-gets-richer regime,” where sub-centers in regions such as Central Europe formed
Figure 1. A still from an animation that was part “A Network Framework of
Cultural History” published in Science Magazine by Maximillian Schich, et al.
Copyright © 2014 Maximillian Schich, University Texas, Dallas.
Author’s Note: I would like to thank the peer-reviewers and editors of Artl@s
Bulletin for their incredibly helpful comments and suggestions.
Maximilian Schich, Chaoming Song, Yong-Yeol Ahn, Alexander Mirsky, Mauro
Martino, Albert László Barabási, and Dirk Helbing. “A Network Framework of
Cultural History,” Science 345, no. 6196 (August 1, 2014): 558–62.
4 The names were collected from Google’s Freebase; Getty’s Union List of Artists
Names; and DeGruyter’s General Artist Lexicon. Ibid, 558.
3
Maximilian Schich, “Figuring Out Art History,” International Journal for Digital Art
History vol. 2 (2016): 40.
2 Ibid, 45.
1
ARTL@S BULLETIN, Vol. 6, Issue 3 (Fall 2017)
4
Visualizing Networks
dispersed patterns as they competed for
dominance.5 Although these findings generally
reproduce common knowledge about the social
structure that governed these regions during the
modern period, Schich argues that pinpointing
specific dates when major cultural hubs in Europe
shift or develop new organizational patterns can
help researchers to locate “historical trends of
cultural centers beyond the scope of specific events
or narrow time intervals.”6
Theorizing a Network Framework
for Art History
The philosopher Michael P. Lynch utilizes Charting
Culture as an example to explore how the
proliferation and positivistic view of data in the
digital age can suppress important theoretical
questions despite the appearance of giving us
greater access to knowledge.10 In regard to Schich’s
video, which draws on the findings of his Science
article, Lynch asserts that the under-examined
questions include: How is culture being defined in
advance of the data analysis? What are the
underlying assumptions that shape the metadata?
How do the sources of the data shape the findings?
What gets left out as a result? And finally, what role
do the actual works of art by these important
figures play in this formulation of culture? 11
Investigating these questions leads to a troubling
set of assumptions that art historians—employing
feminist, Marxist, post-colonial, psychoanalytic,
and semiotic methods—have fought hard over the
past 50 years to overturn. Such assumptions
include: Western Europe’s unrivaled cultural
dominance and history defined by “notable
figures,” otherwise known as the “great man”
theory of history, which suggests known and
noteworthy figures alone transform history. This
theory pushes to the side the power dynamics of
gender, class, race, religion, and ethnicity, while
obscuring social forces such as economics and
politics. In other words, although network analysis
has a great potential to reveal the significance of
actors marginalized by canonical narratives of art
history through the mining of archival data and also
to
track
unforeseen
transnational
and
intercommunal histories of artistic exchange, it
may also paradoxically silence social hierarchies
and mechanisms of marginalization, as well as
historical disruptions to them, if the principles
Schich’s aim with these two articles is to
demonstrate the need for a “systematic science of
art and culture” that “integrates qualitative inquiry
and observation, with methods of computation,
natural science, and information design.”7 For
Schich, this new data-driven approach to art history
is necessary because on the one hand, the
computational ability of digital technology
constitute an under-utilized resources that can help
to reveal unforeseen cultural patterns through the
processing of vast amounts of data; and on the
other, because it could make art history newly
relevant in an economic and political climate in
which the discipline is under threat.8 Certainly
Schich’s video Charting Culture made with
information designer Mauro Martino and produced
by the scientific journal Nature—which currently
has over 1,250,000 views on YouTube—has far
surpassed the reach of the vast major of art
historical publications.9 In our image-saturated
digital environment, projects such as this one that
seek a broader understanding of art history and
visual culture are undoubtedly an important
endeavor. However, one must ask: do the
epistemological premises of this data-driven
approach to culture themselves constitute a threat
to the very humanistic line of inquiry that Schich
argues they’re saving?
Ibid, 562.
Ibid, 558.
Schich, “Figuring Out Art History,” 41-42.
8 Ibid, esp. 42, 45-47.
9 In addition to the large number of views, the video—built on the research of
Schich’s team from the Science article—also sparked an intense, if not always
productive, debate about racial, ethnic, and gender bias of the study in the YouTube
comments section. See: https://www.youtube.com/watch?v=4gIhRkCcD4U
Retrieved August 1, 2017. The video was original published on Nature’s video
channel on July 31, 2014 and was designed by Schich and Martino, along with Kerri
Smith, Charlotte Stoddart, and Alison Abbot:
http://www.nature.com/nature/videoarchive/charting-culture/index.html
10 Michael P. Lynch, The Internet of Us: Knowing more and Understanding Less (New
York: W. W. Norton, 2016), 161. Similarly, media theorist Alexander R. Galloway has
described the positivistic empiricism of computational approaches to analyze culture
(particularly network analysis) as “anti-hermeneutic” or the idea of knowledge
without or against interpretation. David M. Berry and Alexander R. Galloway, “A
Network is a Network is a Network: Reflections on the Computational and the
Societies of Control” Theory, Culture and Society 33, no. 4 (2016): 159.
11 Ibid, 162.
5
6
7
Visualizing Networks
5
ARTL@S BULLETIN, Vol. 6, Issue 3 (Fall 2017)
underlying the data are not interrogated from the
outset.
However, enumerating specific attributes in order
to delineate precisely what constituted a “race film”
was such a thorny endeavor as to be nearly
impossible. Left in somewhat of a quandary, she
asks: “How can a community of practice have no
reliably enumerable qualities, and yet, without a
doubt, constitute a community?”17 For Posner and
her students this un-addressability was the most
fascinating aspect of this project. Although it is easy
to say that the digital humanities constitute an
irresolvable opposition—between an emphasis on
pattern, probability, and consensus in computa-tional analysis, and ambiguity, unpredictability,
and dissensus in the humanities—Posner’s
example demonstrates how working back and forth
between qualitative and quantitative forms of
analysis can help make researchers’ thinking about
a particular subject more systematic or precise,
while at the same time complicate the very process
of classification and quantification.
Schich’s team, however, does acknowledge the
biases of their metadata in the supplementary
materials of their Science article on which Charting
Culture is based, if not in the video itself.12 This
acknowledgement distinguishes Schich’s project
from what visual theorist Johanna Drucker has
described as the problematic “realist” approach to
data analysis and visualization, or the lack of
recognition that “Data are capta, taken not given,
constructed as an interpretation of the phenomenal
world, not inherent in it.”13 In other words, data is
not a neutral or natural phenomenon, but rather an
observer-dependent
construction
that
is
constituted by the epistemologies that shape it.
Moreover, Schich asserts the vital importance of
discussing the structural biases against
underrepresented artists and believes that
quantitative analysis can produce “actionable
insight for future research and funding” to address
these data voids.14 However, even if researchers
worked to create more inclusive databases by
digitizing non-canonical works (which could have
the adverse side-effect of endangering an artwork’s
material survival in more economically precarious
parts of the world),15 one must still ask if the
fundamental premises of quantitative analysis
remain in opposition to humanist thought.
The heuristic value of such a discussion cannot be
emphasized enough. Augmenting established art
historical methods with new digital humanist
techniques can help to broaden the discipline,
particularly when demographics and statistical
maps provide evidence of cultural exchanges that
challenge canonical narratives of art history.
Michele Greet’s Transatlantic Encounters project
on Latin Americans in Paris between the wars is
exemplary in this regard.18 Building and analyzing
large data-sets, Greet and her team offer proof of
the importance of foreign artists to the
development of modern art in France—questioning
the Eurocentrism of current scholarship on
interwar modernism and exploring how a panLatin Americanism began to develop
In my conversation with digital humanist Miriam
Poser at the end of this volume, she expresses her
concern that there is “a terrible problem at the
heart of the digital humanities endeavor: namely,
that the operations necessary to divide sources into
data are in some ways antithetical to the humanistic
enterprise.”16 Illustrating this point with an
example, she describes how she and her students
attempted to create a database of silent race films.
As stated in the article: “Potential sources of bias are addressed in the SM,
including biographical, temporal, and spatial coverage; curated versus crowdsourced data; in- creasing numbers of individuals who are still alive; place
aggregation; location name changes and spelling variants; and effects of data set
language.” Schich, et al. “A Network Framework of Cultural History,” 558.
13 Johanna Drucker, Graphesis: Visual Forms of Knowledge Production (Cambridge,
MA: Harvard, 2014), 128.
14 Schich, Figuring Out Art History, 50.
15 Stephanie Porras makes this point in her article: Stephanie Porras, “Keeping our
eyes open: Visualizing networks and art history,” Artl@s Bulletin vol. 6, no.3 (Fall
2017): http://docs.lib.purdue.edu/artlas/
16 Miriam Kienle, “Digital Art History “Beyond the Digitized Slide Library”: An
Interview with Johanna Drucker and Miriam Posner,” Artl@s Bulletin, vol. 6, no.3 (Fall
2017): http://docs.lib.purdue.edu/artlas/
17 Ibid.
18 Michele Greet, "Mapping Cultural Exchange: Latin American Artists in Paris
between the Wars,” in Circulations in the Global History of Art; Studies in Art
Historiography series, Thomas DaCosta Kaufmann, Catherine Dossin, and Béatrice
Joyeux-Prunel eds. (London: Ashgate: 2015), 133-147. To experience Greet’s
interactive maps, see: Michele Greet, “Transatlantic Encounters: Latin American
Artists in Interwar Paris, http://chnm.gmu.edu/transatlanticencounters/ (Accessed
October 10, 2017). The book that goes along with the online DH project is
forthcoming from Yale, see: Michele Greet, Transatlantic Encounters: Latin American
Artists in Paris between the Wars (New Haven, CT: Yale University Press, 2018),
forthcoming.
12
ARTL@S BULLETIN, Vol. 6, Issue 3 (Fall 2017)
6
Visualizing Networks
Figure 2. A map of Latin American artists in Paris between the World Wars from Michele Greet’s Transatlantic Encounters
project, which is organized by the artists’ nationality, dates in Paris, and birth and death dates. Visitors may explore the proximity
of these artists to each other and to the sites and neighborhoods of Paris during the war, while also examining the works of
individual artists. The website is powered by Omeka (an online exhibition platform) with maps are powered by Google. Copyright
© 2015 Michele Greet, George Mason University, Washington DC.
Visualizing Networks
7
ARTL@S BULLETIN, Vol. 6, Issue 3 (Fall 2017)
among artists living abroad.19 However, the success
of this project stems from how it weaves together
narrative and digital formats in a manner that
integrates uncertainty, subjectivity, and historical
situtatedness into quantitative methods. For
example, although Greet’s interactive maps provide
staggering evidence of the presence of Latin
American artists in Paris during the period—
marking each one by national identity and place of
residence in Paris [Fig. 2]—she also stresses the
importance of including narrative case studies of
artists who resist national categorization (as dualnationals or second-generation immigrants)
and/or who simply reject the idea of “national
identity” altogether.20 Furthermore, rather than
drawing clear links between Latin American actors
in Paris, the network her project articulates is
implied through the concentration of artists in
particular parts of the city at specific moments in
time because, in many cases, there is not
documented evidence of their connection despite
the probability.
In other words, on the one hand, the digital
humanities must come to see how data sets are
structured by epistemologies, and therefore data
analysis cannot precede interpretation; 22 and on
the other hand, the traditional humanities cannot
simply turn its back on the digital—dismissing the
“disambiguation and discreteness” of computa- tional approaches as too antithetical to its basic
tenets of humanism to be addressed. If we do,
society runs the risk of ceding the problems of
interpreting the complexities of our digital
environment to those who do not have sufficient
tools to deal with them.
In Drucker’s recent scholarship, she has attempted
to address the lack of critical engagement with the
explosion of information visualizations, including
those produced by digital humanities projects. As
she describes in her book Graphesis: Visual Forms of
Knowledge Production, the graphical force of
information visualization too often conceals the
premises and parameters on which the data was
constructed; the graphical expression, in other
words, drops back appearing as a passive
presentation (i.e. a “realist” model) rather than an
active qualitative expression of statistical
information (i.e. a “constructivist” model). 23 In the
field of information design, this realist model is
largely the industry standard. The discipline prizes
graphical efficiency, transparency, elegance, clarity,
and directness. As data visualization guru Edward
Tufte famously stated: “Good design is clear
thinking made visible.”24 For Tufte, the best
infographics show the data, not the design—
making complexity coherent in a seamless manner.
25 In these designs, the verbal, visual, and statistical
are all closely integrated in order to reveal the data
through the simplicity of the design; in other words:
“Designs so good that they are invisible.”26
Johanna Drucker, in our interview, has like Posner
and Greet underscored the importance of
integrating
theoretical
questions
and
methodological paradigms of the traditional
humanities with new digital methods, rather than
polarizing the quantitative and qualitative. As she
argues:
Humanistic methods are rooted in the assumption that
epistemologies constitute their objects of inquiry, they
do not simply encounter them as self-evident and
autonomous objects. This means, however, that making
computational models of interpretative activity has to
change approaches to data structures and their
expressions. We are a long way from achieving this, but
the point is to insist on such explorations within the
conception of digital projects, not simply bracket them
out, shrug, and accept computational efficiencies and
expediencies on their own terms of disambiguation and
discreteness.21
Returning to Schich & Martino’s infographic
Charting Culture and evaluating it on Tufte’s terms,
Ibid.
Greet, "Mapping Cultural Exchange,” 141.
21 Ibid.
22 This thought echoes what Alan Liu has described as the problematic inclination of
the digital humanities toward "tabula rasa interpretation” or “the initiation of
interpretation through the hypothesis-free discovery of phenomena." Alan Liu, “The
Meaning of the Digital Humanities,” PMLA vol. 128, no. 2 (March 2013): 414.
23 Drucker, Graphesis, 125-35. Drucker’s idea of a “realist” idea of infographics bares
similarity to what Sean Cubitt has called the “new formalist mimesis” of data
visualization as meta-presentation or the notion that the world is inherently data.
Sean Cubitt, “Data Visualization and the Subject of Political Aesthetics,” in Postdigital
Aesthetics: Art, Computation and Design, eds. David M. Berry and Michael Dieter
(London: Palgrave Macmillan, 2015), 179–81.
24 Edward Tufte quoted in Alan Bisbort, “Escaping Flatland” (1999):
https://www.edwardtufte.com/tufte/advocate_1099, accessed July 20, 2017.
25 Edward Tufte, The Visual Display of Quantitative Information, Second Ed. (Cheshire,
CT: Graphics Press, 2001), esp. 13, 177.
26 Edward Tufte, Envisioning Information (Cheshire, CT: Graphics Press, 1990), 33.
19
20
ARTL@S BULLETIN, Vol. 6, Issue 3 (Fall 2017)
8
Visualizing Networks
one may observe how it succeeds as a transparent,
efficient, and elegant presentation of 2600 years of
cultural history. In fact, Charting Culture was
featured on numerous lists of the “best data
visualizations” of 2014 and 2015.27 However, if we
look more critically, applying the art historical tools
of visual analysis and comparison, we may begin to
see how precisely the graphical display expresses
the data and investigate the broader theoretical
implications of visualizing cultural history in this
manner.
Visual Analysis
Visualization
and
determining “cities of significance,” as these data
points are debatable and cannot as easily appear as
a given. For this reason, experts in macroscopic
data analysis may find it useful to partner with art
historians specializing in particular time periods
and cultures, so that when they zoom in to specific
examples, the microscopic maybe examined in a
manner that provides more nuance, perhaps even
complicating the macroscopic view.
Correspondingly, the processing of large datasets,
as Matthew Lincoln points out in his article, may
help specialists focused on the microscopic to
locate unforeseen continuities, despite the priority
given to disruption-centered narratives within the
discipline of art history. However, in contrast to
Schich’s study, Lincoln’s project only examines
continuities over a few hundred years and it
focuses on his areas of specialization in
Renaissance and Baroque Europe.
Network
In the beginning of the video, we see a blue dot
appear, signifying the birth of Leonardo da Vinci in
the Tuscan village of Anchiano near Florence in
1452 [Fig. 3]. Floating on a white ground, a grey dot
representing Leonardo moves from a blue dot
toward a red one that signifies his death in the
French city of Amboise in 1519. As a fine grey line
arcs across the frame connecting the blue and red
dots, the narrator describes how the “Great
Renaissance Artist” was so well-known across
Europe that late in his life he was invited by the
King of France to come work for him. For those
familiar with Leonardo, the efficient graphic may
appear to omit other key nodes and links (such as
Leonardo’s works for the Duke of Milan or the Pope
in Rome). Perhaps the choice to analyze only one
link (e.g. an artist) between two nodes (e.g. cities of
birth & death) was for the sake of clarity and purity
of the graphic. However, the lack of complexity in
the network diagram may also have to do with the
thorny issue of characterizing other important sites
of production in an artist’s biography. Unlike cities
of birth and death, plotting the city in which an
artist’s “notable works” were made or reside is a
qualitative question that depends on point of view.
Establishing nodes (also termed vertices, actors,
agents, and data points) and links (or arcs, edges,
ties, and relations) is much more difficult when
In Charting Culture, which spans thousands of years
all over the globe, a figure like Leonardo helps to
ground a vast analysis, serving as the paradigmatic
example for all the other actors and relations. With
the thousands of agents and their ties culled from
Google’s Freebase.com, a world map of global
culture unfolds [Fig. 4]. While the narrator
describes Schich’s research project, arcs of a
gradient blue/red hue spread across the flat grey
world map—first in Europe and then elsewhere—
as hundreds of years accumulate like seconds on a
small stopwatch in the upper left-hand corner.
Covered in nodes and links, this animated map
closely resembles graphics that visualize global
flight patterns and internet connectivity—
diagrams to which we have all become accustomed
in the 21st century. Unlike the timelines that have
traditionally anchored historical texts, this new
“network framework” of cultural history prioritizes
space over time.
For three of these lists, see: Nathan Yau, “Best Data Visualizations in 2014”
FlowingData: http://flowingdata.com/2014/12/19/the-best-data-visualizationprojects-of-2014-2/ Retrieved August 1, 2017; Gareth Cook, ed. Best American
Infographics 2015 (New York: Houghton Mifflin Harcourt, 2015), 150-51; National
Science Foundation & Popular Science, 2015 Vizzies Visualization Challenge - Finalist
- Video: http://www.mamartino.com/img/nsf_award.jpg Retrieved August 1, 2017
27
Visualizing Networks
9
ARTL@S BULLETIN, Vol. 6, Issue 3 (Fall 2017)
Figure 3. A still that centers on the life of Leonardo da Vinci from the video Charting Culture made by professor Maximillian Schich (University Texas, Dallas) and
designer Mauro Martino (IBM Watson Group) with Kerry Smith, Charlotte Stoddart, and Alison Abbott at Nature. Copyright © 2014 by Nature/Macmillan Publishers
Ltd.
Figure 4. A still that visualizes global cultural migration from the video Charting Culture (2014) made by professor Maximillian Schich (University Texas, Dallas) and
designer Mauro Martino (IBM Watson Group) with Kerry Smith, Charlotte Stoddart, and Alison Abbott at Nature. Copyright © 2014 by Nature/Macmillan Publishers
Ltd.
ARTL@S BULLETIN, Vol. 6, Issue 3 (Fall 2017)
10
Visualizing Networks
This shift from a temporal to spatial emphasis,
however, is not unique to the Charting Culture
project. Many network visualizations and statistical
maps minimize the complexity of temporality and
the importance of periodization in profit of a stable
geography. Although there are practical and
technical reasons as to why this might be the case,
28 the prioritizing of geography over chronology
may also be seen as a structuring principle of what
the sociologist Manuel Castells calls the “network
society,” or a society in which “space organizes
time” and space is presented as fluid rather than
bounded.29 Castells argues that under the
conditions of globalization, connectivity controlled
by a global elite has created a society in which the
“space of flows dominates the space of places” and
“timeless time supersedes clock time.” 30 Although
historical work has always entailed looking at the
past
with
present-day
frameworks
of
understanding, the network frameworks of our
digital age may imperil the concrete time of specific
localities by collapsing it into a “timeless time” of
global connectivity that works against historical
thinking.31
groups.33 As Stephanie Porras argues in her article,
we must understand network visualizations of
historical
phenomenon
as
“artifacts
of
contemporary visual culture, laden with the biases
and limits of both past and present knowledge
systems.”34 In the instance of Charting Culture, the
bias of Eurocentric databases—a legacy of the
enforced inequities of European colonial
domination—are clearly visualized in the far larger
concentration of data points in the global north.
However, against this implicit colonialist discourse,
Schich and Martino have chosen not to use the
standard world map or the “Mercator Projection”—
used by Google Maps, among others—which has
been long criticized for how it grossly minimizes
the actual size of Africa and Latin America. In
Charting Culture, the viewer gazes from South to
North which offsets the distortion. And yet, due to
the admitted bias of the data, when the camera pans
and zooms in on Japan or Australia, the names that
inhabit these spaces are predominately European
rather than Japanese or Aboriginal.35 The
discrepancy visualized here, as Schich later
suggested, can be seen as a strong argument for
larger and more comprehensive digitizing projects
that are aimed at addressing such imbalances. 36
Against the pejorative manner in which “big data”
is discussed by many art historians, Schich argues
that we in fact need “bigger data,” in order more
holistically and critically examine cultural
exchange.37 Furthermore, as Léa Saint-Raymond &
Antoine Courtin assert in their contribution to this
issue, linked open data may also be utilized to
All network visualizations, however, do not
function in the same manner and thus
demand critical analysis. There are “countercartographies,” such as the one Greet constructed,
which was researched and designed from the
perspective of a scholar trained in post-colonial
studies.32 As the field of critical cartography has
taught us, we must examine the assumptions that
structure the network diagram, as neither the map
nor the data that shapes it are neutral, and often
they serve and reflect the interests of dominant
For example, Michelle Greet observed that her team faced challenges when
attempting to integrate time into maps in their Transatlantic Encounters project. This
was in part due to the lack of, or discrepancies between, the dates of the artists’
residential data. Without chronology that shows changes in artists residences over
time, as Greet admits, their map “creates a false sense of contiguity.” The challenge of
adequately time also had to do with the choice of platform (i.e. a geo-coded
GoogleMap). Greet, "Mapping Cultural Exchange,” 142.
29 Manuel Castells, The Rise of the Network Society, second edition (Oxford: Blackwell,
2000), 407. I have also shown how this move out of time and into space works in
other art historical network analyses: Miriam Kienle, “Artist as Network,” MediaN, vol. 11, no. 2 (Fall 2014):
http://median.newmediacaucus.org/art-infrastructures-information/artist-asnetwork-out-of-time-and-into-space-with-the-warholtimeweb
30 As Castells observes of the network society: “Space and Time, the material
foundations of human experience, have been transformed, as the space of flows
dominates the space of places, and timeless time supersedes clock time of the
industrial era.” Manuel Castells, End of Millennium: The Information Age: Economy,
Society, and Culture, Volume 3. Second Edition (Malden, MA: Wiley-Blackwell, 2010), 1.
Castells, Network Society, 1-25, esp. 5-12.
Greet, "Mapping Cultural Exchange,” 133-37.
For an excellent overview of the field of critical cartography, including examples of
“counter-cartography” that express the worldview of non-dominant groups, see:
Jeremy W. Crampton and John Krygier, “An Introduction to Critical Cartography,”
ACME: An International Journal for Critical Geographies vol. 4, no. 1 (2006): 11–33.
On current issues in the field, see the special issue of this journal on the subject “Do
maps lie?”: Béatrice Joyeux-Prunel, ed. “Introduction: Do maps lie?,” Artl@s Bulletin
vol. 2, no. 2 (2013): http://docs.lib.purdue.edu/artlas/vol2/iss2/
34 Porras, “Keeping Our Eyes Open,” http://docs.lib.purdue.edu/artlas/
35 The notable figures highlighted by the narrator are 17 th century Christian
missionaries to Japan who were “persecuted for their faith” by the Japanese, and the
infamous “bushranger” Ned Kelly, the son of an Irishman (i.e. a British colonial
subject) banished to Australia by the British because he stole pigs. The narrative
therefore is not only Euro-centric in the selection of individual figures, but also in the
dichotomous savage/civilized narrative of colonialism that it reinforces.
36 Schich, “Figuring Out Art History,” 50.
37Ibid, 47-51.
28
Visualizing Networks
31
32
33
11
ARTL@S BULLETIN, Vol. 6, Issue 3 (Fall 2017)
Figure 5. Four stills that together show the “pull of the west coast cultural centers” in US during the 19th and 20th centuries from the video Charting Culture (2014)
made by professor Maximillian Schich (University Texas, Dallas) and designer Mauro Martino (IBM Watson Group) with Kerry Smith, Charlotte Stoddart, and Alison
Abbott at Nature.. Copyright © 2014 by Nature/Macmillan Publishers Ltd.
address the publication bias of any one digital
archive alone.38 With these promising new
developments on the horizon, one might wonder:
How could bigger data that is open and linked
transform a visualization like Charting Culture?
With the emergence of new data points, what kind
of new narrative might surface?
there,” particularly migrating toward cultural and
technological meccas like San Francisco and Los
Angeles [Fig. 5A-5D].
Ultimately, we are told: “This depiction makes
clearer the pull of the west coast cultural centers
where people move during their lifetimes.” Similar
to the rhetoric that underpinned the US’s westward
expansion (i.e. manifest destiny), the “pull of the
west” appears as a natural phenomenon, and thus
mythical rather than historical.39 However, perhaps
by accumulating more data related to the cultural
contributions of indigenous populations, the push
and pull of creative actors might look quite
different.
Turning to US cultural history (after an analysis of
the development of major European cultural
centers from Roman times to present), the video
quickly pans to actors flocking from Europe to the
northeast coast of the United States, and then out to
the west coast from the 18th century onward. As
these cultural migrants flow across the country, the
narrator describes how at first notable people
settled in the river valleys, but then with the advent
of new “transportation links” larger numbers
flocked to the west coast to “live out their days
Watching this massive migration from the
perspective of a scholar of American Art, it is hard
not to wonder about the people who were already
there; how their culture intersected with that of the
Lea Saint-Raymond & Antoine Courtin, “Enriching and Cutting: How to Visualize
Networks Thanks to Linked Open Data Platforms,” Artl@s Bulletin vol. 6, no.3 (Fall
2017): http://docs.lib.purdue.edu/artlas/
39 Describing the force of technological progress and westward expansion—while
suppressing the violent colonization of Native American lands, and the attempted
erasure of their people and culture—Charting Culture like manifest destiny makes
the Euro-American domination of the West appear to be a natural, divine, or
inevitable phenomenon, rather than the product of ideology and the policies and
archives constituted by it. In this way, the video is “mythical” rather than “historical”
in Roland Barthes’ sense of the term because it transforms “history into nature”
rather than investigating its relation to dominant period ideologies. Roland Barthes,
“Myth Today,” in Jessica Evans & Stuart Hall, ed. Visual Culture: The Reader (London:
Sage, 1999), 58.
38
ARTL@S BULLETIN, Vol. 6, Issue 3 (Fall 2017)
12
Visualizing Networks
settlers; and where and how they “lived out their
days.” During a brief pause in the narrative, we
zoom in on the west coast around 1850 to see a
flood of European male names attached to nodes
that float freely across the landscape [Fig. 5B].
Looking hard among these names, one may find a
few indigenous names, such as Cochise (a leader of
the Chiricahua Apache) and Sem-Yeto (also known
as Chief Solano, a famous leader of the Suisunes).
However, in contrast to the colonizers, these actors
appear more static, resting in a single spot for the
duration of the clip, despite the historical reality of
being forced to leave their native lands during
periods of US national expansion.40 Perhaps
collaborating with art historians trained in Native
American studies and postcolonial methods, the
migrations of indigenous artists could become
more accurately documented and narrative
accounts of the contributions of these figures would
emerge (such as the one about Leonardo Da Vinci)
that could help to begin to address the video’s bias.
perverse capacity—honed to perfection in the
history of science tied to militarism, capitalism,
colonialism, and male supremacy—to distance the
knowing subject from everybody and everything in
the interests of unfettered power.” 42 The myth of
this neutral, distanced, and all-encompassing form
of vision becomes a powerful tool that suppresses
that fact that all vision is embodied and knowledge
socially situated. In order to humanize and socially
situate cultural production and reception, we must
therefore—as the next section argues—move back
and forth from big to small, quick to slow, global to
local, always acknowledging the assumptions that
frame those movements.
Investigating Abstraction: Intimacy,
Distance, and Para-Empiricism
As many network visualizations employ this
distanced macroscopic perspective, it is important
to investigate how it operates. One particularly
notable example in the field of art history is the
“Artist Network Diagram” from the Museum of
Modern Art’s Inventing Abstraction exhibition
(December 23, 2012-April 15, 2013) [Fig. 6].
Lauded initially by critics as presenting a new, more
inclusive and interdisciplinary approach to the
history of artistic abstraction,43 recent scholarship
on the interactive network diagram has addressed
the implications of viewing this history
at a
remove—as an abstraction of the
Finally, as we zoom out again to view the entire
country, the narrator describes how the invention
of trains, and then cars, facilitated the westward
migration of notable figures [Fig. 5C]. Evocative of
airplane networks, brilliantly colored arcs connect
cities and towns across the land—grafting the
image of present-day transportation technologies
onto those from the past. The presentism and
techno-positivism of the video are furthered by the
fantasy of being able to see the whole world at once,
or what feminist philosopher and new media
theorist Donna Haraway famously called the “god
trick” of “seeing everything from nowhere.” 41 The
totalizing vision offered by visualization
technologies, according to Haraway, have “a
As the second half of the 19th century has been characterized by Michael Friendly
as the “golden age of statistical graphics,” it would be interesting to contemplate the
role that infographics played in the US’s Native American reservation system. In his
book Representing the Nation, John Rennie Short has argued that the great American
statistical cartographer Francis A. Walker famously combined census data into
thematic maps in his widely-read Statistical Atlas (1874), which was aimed at
national unification after the civil war as well as social reform and control. As a
proponent of the reservation system, Walker’s emphasis on westward expansion
worked hand-in-hand with his focus on locating spaces of social disorder or
deviation from the colonizer’s norm. Connecting race to spaces of moral and
physical degeneracy through data, Walker aimed to enumerate the “other.” The
visualization of census data was thus used to justify discriminatory policies, as well
as the continued accumulation and visualization of census data through visually
striking designs. Michael Friendly, “A Brief History of Data Visualization,” Handbook
of Data Visualization (Berlin: Springer-Verlag, 2008), 28. Charles Rennie Short,
Representing the Republic: Mapping the United Stated, 1600-1900 (London: Reaktion,
2001), 217-20.
Donna Haraway, “Situated Knowledges: The Science Question in Feminism and the
Privilege of Partial Perspective,” Feminist Studies vol. 14, no. 3 (1988): 575–599,
quote 581.
42 Ibid.
43 For positive reviews, see: Roberta Smith, “When the Future Became Now,” The
New York Times, December 20, 2012,
http://www.nytimes.com/2012/12/21/arts/design/inventing-abstraction-19101925-at-moma.html (accessed September 15, 2017); Thomas Micchelli, “MoMA’s
Show of Shows: ‘Inventing Abstraction, 1910–1925,’” Hyperallergic, December 22,
2012,
https://hyperallergic.com/62402/momas-show-of-shows-inventing-abstraction1910-1925/ (accessed September 15, 2017); Peter Schjeldahl, “Shape of Things: The
Birth of Abstraction,” The New Yorker, January 7, 2013,
http://www.newyorker.com/arts/critics/artworld/2013/01/07/130107craw_artw
orld_schjeldahl (accessed September 15, 2017); Hal Foster, “At MoMA,” London
Review of Books, February 7, 2013, https://www.lrb.co.uk/v35/n03/hal-foster/atmoma (accessed September 15, 2017).
40
Visualizing Networks
41
13
ARTL@S BULLETIN, Vol. 6, Issue 3 (Fall 2017)
Figure 6. The Artist Network Diagram in Inventing Abstraction: 1910-1925, an exhibition at The Museum of Modern Art, 2012-13, organized by Leah Dickerman with Masha
Chlenova. Courtesy of the Museum of Modern Art. Copyright © 2012 The Museum of Modern Art, New York.
mechanisms of social connectivity.44 In this
interactive network visualization, the user is
prompted to explore the connections between
important pioneers of abstraction—from Georgia
O’Keeffe and Alfred Stieglitz in the US, to Pablo
Picasso and Sonya Delaunay in France, to Natalia
Goncharova and Mikhail Larinov in Russia—who
are all linked by red lines that spread across a plain
cream background (despite the configuration
implying the geography of the US and Europe).
Artists with the greatest connectivity, such as
Delaunay, Goncharova, and Picasso are highlighted.
With this densely connected web of artists, the
visualization productively shows that abstraction
was not an innovation of “solitary genius,” but a
product of “network thinking.” It also demonstrates
that female artists (traditionally marginalized in
the history of abstraction) were in fact central hubs
in the development of this revolutionary art form.45
In fact, revealing the significance of important yet
under-recognized female artists—through analysis
of archived correspondence, institutional records,
and auction results—is one manner in which
network analysis in the field of art history can be
particularly useful from a feminist perspective. 46
For more recent critiques, see: Miriam Posner, “MoMA’s Inventing Abstraction”
Network Diagram” DH101 Blog, November 9, 2014,
http://miriamposner.com/dh101f14/?p=938; Claire Bishop, “Against Digital Art
History,” Humanities Futures, October 20, 2015,
https://humanitiesfutures.org/papers/digital-art-history/, Maximilian Schich,
“Figuring Out Art History,” 52-3, 60; Nicole E. Reiner and Jonathan Patkowski,
“Inventing Abstraction, Reinventing Our Selves: The Museum of Modern Art’s Artist
Network Diagram and the Culture of Capitalism,” Rutgers Art Review vol. 32 (2017),
http://rar.rutgers.edu/current-volume/volume-32/reiner_patkowski_inventingabstraction-reinventing-our-selves/#go%20to%20notes
45
Leah Dickerman, ed. Inventing Abstraction (New York: Museum of Modern Art,
2013):
https://www.moma.org/interactives/exhibitions/2012/inventingabstraction/
(accessed September 15, 2017).
46 On the benefits of using network analysis to recover important women artists and
their networks, see: Christina Weyl, “Networks of Abstraction: Postwar Printmaking
and Women Artists of Atelier 17,” Archives of American Art website (Graduate
Student essay prize, 2014): http://www.aaa.si.edu/essay/christina-weyl; On the
problems and promises of network analysis from a feminist perspective, see:
Michelle Morevec’s contribution to this volume: "Network Analysis and Feminist
Artists," Miriam Kienle, ed. “Visualizing Networks” Special Issue, Artl@s Bulletin 6, no
.3 (Fall 2017): http://docs.lib.purdue.edu/artlas/
44
ARTL@S BULLETIN, Vol. 6, Issue 3 (Fall 2017)
14
Visualizing Networks
However, as feminist digital humanists Catherine
D’Ignazio and Laura Klein have pointed out,
feminist data visualization entails examining
power, situating knowledge, and making one’s
process visible.47 Therefore, visualizations that
reinforce masculinist ideals of distance, mastery,
and totality, as well as hierarchical and binaristic
configurations like center/periphery, cannot be
called “feminist” per se.48 Therefore, although
MoMA’s network diagram appears to disrupt
hierarchies through the horizontal placement of
nodes and demonstration of documented patterns
of acquaintance that reveal women as key players,
one is still left wondering what exists beyond this
Euro-and-US-centric web of names floating on a
blank background. As a result of the design, it is
hard to situate these agents in space and time and
to understand the power that the highlighted actors
are meant to hold. As Miriam Poser points out in her
analysis of this network visualization, despite being
told that “24 connections” serves as a threshold for
an artist to be highlighted, it is uncertain what that
criteria signifies (influence, inclusion, or
extroversion) or from where these connections are
derived (letters, magazines, or exhibition
catalogues).49 Similarly, Maximillian Schich has
critiqued the diagram for its lack of complexity and
precision. For Schich, the famous 1936 flow chart
from Cubism and Abstract Art by MoMA curator
Alfred Barr—which the Inventing Abstraction
diagram is meant to update [Fig. 7]—has far more
nuance and art historical value.50
artists; 3) the reduction of artistic labor and
innovation to an entrepreneurial model of social
connectivity that resonates with a contemporary
neo-liberal world-view.51
Figure 7. Chart prepared by Alfred H. Barr, Jr., for the dust jacket of the
exhibition catalogue, Cubism and Abstract Art, by Alfred H. Barr, Jr. Copyright
© 1936 The Museum of Modern Art; Licensed by SCALA / Art Resource, NY.
Building on Reiner and Patkowski’s argument, the
art historian Claire Bishop states: “Carefully
reasoned historical narrative is replaced by social
network (the avant-garde equivalent of LinkedIn)
and has no room for non-human agents that elude
quantification—such as African artifacts, which
were crucial to the development of abstraction, or
the imperial powers that mobilized their
circulation in Europe.”52 Developments in art and
culture thus appear to emerge from free and even
flow of ideas among singularly significant actors,
rather through a process of exchange among known
In Nicole Reiner and Jonathan Patkowski’s recent
analysis of the Inventing Abstraction network
visualization, they too found several troubling
implications: 1) the omission of non-western
cultures, which the original Barr flow chart (while
problematic in other ways) includes and even
highlights; 2) the elision of powerful economic and
social forces that shape the movement of art and
Catherine D'Ignazio and Laura Klein, “Feminist Data Visualization,” IEEE VIS 2016
Workshop on Visualization (unpublished conference paper),
http://vis4dh.dbvis.de/papers/Feminist%20Data%20Visualization.pdf (accessed
September 15, 2017). Catherine D'Ignazio, “What would feminist data visualization
look like?,” MIT Center for Civic Media Blog, December 20, 2015,
https://civic.mit.edu/feminist-data-visualization (accessed September 15, 2017)
48 Ibid.
Posner, “MoMA’s Inventing Abstraction”
http://miriamposner.com/dh101f14/?p=938
50 Schich, “Figuring Out,” 52-3, 60.
51 Reiner and Jonathan Patkowski, “Inventing Abstraction, Reinventing Our Selves”
http://rar.rutgers.edu/current-volume/volume-32/reiner_patkowski_inventingabstraction-reinventing-our-selves/#go%20to%20notes
52 Claire Bishop, “Against Digital Art History,”
https://humanitiesfutures.org/papers/digital-art-history/
47
Visualizing Networks
49
15
ARTL@S BULLETIN, Vol. 6, Issue 3 (Fall 2017)
and unknown agents that entails uprooting, loss,
intersection, acquisition, and repurposing of
cultural forms or what is known in the field of visual
studies as “transculture.”53
By replacing single paradigmatic examples (i.e. the
case study approach of traditional art history) with
a dizzying array of loosely related images that
recalls the results of a Google search, Bishop aims
to critically address the pervasiveness of distance
viewing in contemporary digital culture. As Bishop
notes in her article, this distanced and
decontextualized mode of image viewing has
infuriated audience members, which is perhaps
precisely the reaction she aims to achieve. A
believer in the necessity of antagonism to
democracy, Bishop sees this mode of critique as a
catalyst for change.56 However, one must wonder, if
no alternative is proposed—no suggestion as to
where to go from here—doesn’t this presentation
of our digital image landscape simply reinforce the
numbing effects it’s meant to critique?
Bishop’s critique of the Inventing Abstraction
diagram, however, is not simply skeptical of
network frameworks of art and culture, but of
digital art history more broadly. Bishop—whose
current research analyzes the impact of digital
technology on global contemporary art—has
emerged as a particularly vocal critic of digital art
history. In a recent paper, she suggests that digital
or quantitative approaches to art history are also
the product of the neo-liberalism, which has
transformed all activity into metric evaluation,
even if not directly monetized. As she states:
Digital art history, as the belated tail end of the
digital humanities, signals a change in the character
of knowledge and learning. Ideals like public
service, citizenship, knowledge as an end in itself,
and questions of what is just, right, and true have
decreasing validity because they resist quantitative
measurement, and moreover do not easily
translate into information that optimizes the
performance of society (i.e. generate profit).54
Against such polarizing, the architect and visual
studies scholar, Laura Kurgan, has taken a different
tack. In her book Up Close At A Distance, she not only
attempts to critique, but also to expand and
improve how digital technologies of representation
are utilized. These projects, as Kurgan explains,
“explicitly reject the ideology, the stance, and the
security of 'critical distance' and reflect a basic
operational commitment to a practice that explores
spatial data and processing from within." 57 As
founder of Columbia University’s Spatial
Information Design Lab (SIDL) and director of the
Center for Spatial Research (CSR), she and her team
examine the political and ethical stakes of digital
technologies—working with and through them to
gain intimate knowledge of how these tools may be
utilized to create greater civic engagement and a
more just and democratic world. For example, in
recent visualizations of the global migrations that
stem from political, economic, and environmental
crises such as EXIT (2015) and Conflict Urbanism:
Columbia (2016) [Figs. 8 & 9], Kurgan
acknowledges the military and governmental
origins of the technologies and modes of
visualization she employs to examine these spaces
In Bishop’s view, the neo-liberalization of higher
education in the US academy has worked hand-inhand with the digital humanities to displace (rather
than expand) the public humanities and its
qualitative questions of truth, justice, and liberty.
Furthermore, as art history belatedly adopts the
“distant viewing” of the digital humanities (wherein
vast numbers of catalogued artifacts are examined
through quantitative computational analysis),55
Bishops suggests appropriating this mode of
viewing against itself. Therefore, in presentations
of her current research project, "Déjà Vu:
Reformatting Modernist Architecture," she engages
distant viewing by setting her PowerPoint to
quickly scroll through various combinations of
hundreds of examples of contemporary art from all
over the world that quote modernist architecture.
On transculture, see: Nicholas Mirzoeff, “Transculture: From Kongo to Congo,” An
Introduction to Visual Culture (London: Routledge, 1999): 125-156.
54 Ibid.
55 Bishops idea of “distant viewing” builds on Franco Moretti’s term “distant reading”
or the quantitative approach to comparative literature. Ibid.
Claire Bishop, “Antagonism and Relational Aesthetics,” October no.110 (Fall 2004):
51-80, esp. 65-70; Artificial Hells: Participatory Art and the Politics of
Spectatorship (London: Verso, 2012).
57 Laura Kurgan, Up Close At A Distance: Mapping, Technology, and Politics (Brooklyn,
NY: Zone Books; MIT Press, 2013), 14.
53
ARTL@S BULLETIN, Vol. 6, Issue 3 (Fall 2017)
56
16
Visualizing Networks
and events—traces of which remain even as her
team aims to redirect their utilization towards
humanistic and public goals.
However, the team attempts to de-stabilize the
techno-positivism of statistical cartography and
social network analysis. Rather than offer viewers a
secure and solid orientation, the team’s
visualizations favor complexity and disorientation.
Zooming in and out of specific localities, these
projects embody both a top-down & bottom-up
account of global conflict and suggest that through
a combination of case studies and macroscopic
analysis we might begin to see how new networks
could emerge from this crippled infrastructure. As
Kurgan states, they did not aim to visualize “hard
data, ” but rather to create a “soft map that is
infinitely scalable, absolutely contingent, and open
to vision thus revision.”58
In order to achieve this aim, Kurgan has adopted a
“para-empirical” approach, in which data is treated
not as purely empirical, concrete, and irreducible
facts about the world, but rather a representation
or emissary of reality.59 From this view, data works
alongside reality—remediating and reshaping it—
and it does so incompletely and imperfectly, as
every representation is an interpretation,
negotiation, and translation. Para-empiricism is
thus not meant to dismiss objectivity—plunging us
into an abyss of doubt—but rather open up political
and ethical debates that enlarge our conception of
the world. Although Kurgan works predominantly
in the disciplines of architecture and urban
planning, we might ask ourselves how this “paraempiricism” could inform digital approaches to art
history such as network analysis. If data is not
understood as objective truth—but rather a
representation
that
implies
remediation,
incompletion, and construction—perhaps trans-cultural exchange may be mapped and charted in
new ways. By critically examining the abstraction of
lived experience that constitutes network analysis
and visualization, we can investigate what lies
between all of those nodes and edges, structuring it
in absence.
Figure 8. This digital environment visualizes the 19.5 million
refugees worldwide and more than 38 million IDPs. The team
describes of the project’s aims: “Exit, immerses the viewer in a
dynamic presentation of data documenting contemporary human
movement. Statistics documenting population shifts are not
always neutral and the multiple efforts to collect them
are decentralized and incomplete. Here the data are
repurposed to build a narrative about global migration and its
causes. The viewer enters a circular room and is surrounded by
a panoramic video projection of a globe which rolls around the
room printing maps as it spins. The maps are made from data
which has been collected from a variety of sources, geocoded,
statistically analyzed, re-processed through multiple
programming languages and translated visually. The presentation
is divided into narratives concerning population shifts,
remittances, political refugees, natural disaster and sea-level rise
and endangered languages.” Originally completed in 2008, EXIT
has been fully updated to coincide with Cop21, the United Nations
Conference on Climate Change and reflects data from 2015. On
view at the Palais Tokyo in Paris from November 25, 2015 –
January 10, 2016. Project team © Diller Scofidio + Renfro, Mark
Hansen, Laura Kurgan, and Ben Rubin, in collaboration with
Robert Gerard Pietrusko and Stewart Smith. Photos © 2015 Luc
Boegly.
58
Ibid, 204.
Visualizing Networks
59
17
Ibid, 34-36.
ARTL@S BULLETIN, Vol. 6, Issue 3 (Fall 2017)
analysis, particularly if truly integrated with
computational methods rather than treated as an
afterthought. Frequently designers and computer
scientists are brought into digital humanities
projects late in the game, seen as providing “tools,”
and thus the capacities for digital methods to
investigate and visualize a particular corpus are illor-under-utilized. Additionally, theorizing must be
positioned not as something to be overcome by
quantifiable data, but rather as a fundamental
means of understanding how and why the data is
structured in one way rather than another.
Therefore, the contributors to this volume attempt
to locate more holistic approaches to network
analysis in art history, investigating the
epistemologies that structure a given network and
the knowledge produced in their visualization.
In this issue, Matthew Lincoln’s contribution
demonstrates how computational analysis of
professional
networks
among
European
printmaking communities from the mid-16th to the
mid-18th centuries can expose patterns of exchange
beyond the individualist and nationalist narratives
of artistic development that have traditionally
characterized the histories of this work. Using data
culled from the databases of prints in the British
Museum and Rijksmuseum, Lincoln aims to test
post-hoc national classifications of prints in their
collections rather than assuming them as a
foundational truth. Using network analysis to trace
associations among European communities of print
designers, plate cutters, and publishers, Lincoln
demonstrates how printmaking internationalized
during this period and examines this trend in
relation to powerful demographic forces more than
discrete historical ruptures. Lincoln’s method
recuperates the historical continuities lost in the
disruption-focused narratives that center on the
contributions of exceptional individuals and
anomalous events, showing gradual and
continuous changes can provoke new questions
about the production and circulation of prints
across Europe during the long 17th century,
including the examination of important period
Figure 9. Visualization of Colombia’s internally-displaced people. Data from
Registro Único de Victimas—Red Nacional de Información, Columbia 19852015. The Conflict Urbanism: Colombia (2016) project by Laura Kurgan, Juan
Saldarriaga, and Angelika Rettberg at Center for Spatial Research at Columbia
University, along with M.A. Program in Peacebuilding at Universidad de los
Andes in Bogotá. Copyright © 2016 Center for Spatial Research, Columbia
University, New York.
Analyzing Networks in Art History
“Visualizing Networks: Approaches to Network
Analysis in Art History” features case studies that
utilize network theory and analysis to investigate
patterns of art historical circulation and exchange.
It aims to address the lack of dialogue and
understanding
between
quantitative
and
qualitative approaches to art history, which are too
often polarized. By bringing together articles that
conceptualize art historical networks in different
ways, this issue attempts to ask questions about the
promises and problems of network analysis in the
field of art history. Contributors, therefore,
investigate
how
theoretical
methods
of
humanists—such as feminist and postcolonial art
history—may be productively used in conjunction
with quantitative approaches to art historical
ARTL@S BULLETIN, Vol. 6, Issue 3 (Fall 2017)
18
Visualizing Networks
figures obscured by canonical narratives of art
history.
Yael Rice’s contribution shows how social network
analysis (SNA) of artists in the royal atelier of the
Mughal court during the late 16 th century can help
us better understand both the social structure of
the manuscript workshop (which is little
documented in the period texts) and the stylistic
innovations found in the manuscripts it produced.
Although the unique illustrative style of these
Mughal manuscripts has been framed from a
Eurocentric perspective as the result of the
influence of European prints, or from a pan-Asianist
perspective as the product of a synthesis between
Persianate and Indic cultural systems, Rice shifts
our focus to show how the internal structure of the
manuscript workshop itself played a role in
generating the exceptional commixture of styles
found in these works. Because research on these
manuscripts often focuses on “master” artists,
scholars have not adequately analyzed the
significance of the collaborations and large number
of connections that the workshop actually fostered.
Employing Granovetter’s theory about the
“strength of weak ties,” Rice demonstrates how
Mughal workshops in South Asia fostered
numerous acquaintanceships between artists
(rather than a few intimate or familial
relationships), which resulted in wide stylistic
fusion and highly productive workshops. In order
to arrive at this conclusion, Rice examined the
collaboration between designers and colorists
documented in contemporaneous marginal
inscriptions that appear in three heavily illustrated
royal manuscripts. With SNA software Gephi, Rice
was able to track patterns of collaboration.
However, as Rice admits, the task of gathering,
editing, and cleaning her data in advance of its
analysis in Gephi presented many challenges, such
as spelling inconsistencies of artist names and
shifting roles in the workshop. Rather than elide
these issues, Rice is transparent about the
constructedness and imperfection of her data and
the reasoning behind how and why she edited it in
a particular manner.
Stephanie Porras, who like Lincoln works on
Renaissance and Baroque art and print culture,
examines how networks visualizations have on the
one hand the ability to turn messy archival material
into powerful webs of connection that may reveal
hidden historical actors, and yet on the other hand,
also have the propensity to obscure the biases and
assumptions that structure the archives from which
the network visualization’s data derives. Although
Porras utilizes theories of networks in her research
on the global circulation, reception, and
remediation of a sixteenth-century Flemish print of
St. Michael the Archangel—specifically sociological
theories of “virality” that stress the agency of all
actors in the network and the power structures that
propel movement—she is skeptical about network
science’s ability to analyze art historical patterns
given the structural imbalances that constitute the
available data.60 As she describes, networks require
data; however, data related to art by historically
marginalized populations is often scant,
particularly if the producer was anonymous.
Therefore, network visualizations have the
dangerous potential to “simply re-inscribe historic
and contemporary power differentials—between
colonizer and colonized, places and people that
benefitted from political, economic and social
stability that enabled the accumulation of records
and archives and those that did not.” Porras
demonstrates how this “computational inequality”
plays out in a number of humanist network
visualization projects. Although she recognizes the
formidable task of analyzing vast stores of data and
the potential of computational analysis to help us
better understand the transitivity and centrality of
certain art historical agents and spheres, she also
asserts the need to “keep our eyes open” to the
under-examined assumptions that structure
networks and the biases that their visualization
may unwittingly reinforce.
For Porras’s use of theories of “virality” to understand early modern print
networks, see: Stephanie Porras, “Going Viral? Maerten de Vos’ St. Michael the
Archangel,” Nederlands Kunsthistorisch Jaarboek 66 (2016): 54–78.
60
Visualizing Networks
19
ARTL@S BULLETIN, Vol. 6, Issue 3 (Fall 2017)
Similarly, Michelle Moravec addresses some of the
structural and systemic barriers she faced when
employing social network analysis to examine
feminist artists’ networks. She looks at two
different projects in which she used computational
modes of analysis to study the shape and function
of social networks in the field of feminist art. The
first is a study of the artist Carolee Schneemann’s
social network. Although several recent digital art
history projects—such as the Warhol TimeWeb, Six
Degrees of Peggy Bacon, and Whistler and Roussel:
Linked Visions61—have similarly aimed to expand
the monographic approach to artists by
demonstrating the significance of creative
collaborations,
Moravec
utilizes
network
visualization not simply to display the importance
of networks, but to investigate the power that its
form exerts upon the actors and edges that
comprise it. Although every visualization is an
interpretation on some level, graphical displays
often conceal the epistemologies that structure
them. Moravec’s study, however, is generative and
clearly from a feminist perspective. As a result, in
her analysis of Schneemann’s network, she actively
limits her dataset to female correspondents in
order to see and analyze connections between key
women who are otherwise obscured by the
preponderance of male colleagues, indicative of the
patriarchal structure of the art world. In her second
project, she examines a corpus of American
feminist periodicals to trace the circulation of
feminist art manifestos. With both projects,
Moravec realizes her research was limited: 1) by
the lack of feminist artists’ archives and limited
data pertaining to them 2) by the use of only one
form of computational analysis. She therefore
strived to expand and edit her corpus in ways that
addressed
gender
bias,
and
combined
computational linguistics and network analysis to
create a more nuanced picture of the cultural
community under analysis.
As digital humanist Elijah Meeks has described, the
forms of computational analysis most used by
scholars in the humanities or the “three pillars to
DH research” are “Text Analysis, Spatial Analysis
and Network Analysis,” adding that he also has “a
sneaking suspicion that Image Analysis is
something else that sits with the aforementioned
three.”62 While some digital humanists use one
mode of analysis, many use multiple. Just as
Moravec uses textual analysis alongside social
network analysis, and Maximillian Schich couples
network analysis with spatial analysis, the
contribution by Stefanie Schneider and Hubertus
Kohle makes use of both image and network
analysis. Although in the field of art history the
exploration of sociocultural networks is often done
in conjunction with cartographic analysis,
Schneider and Kohle point out that networks need
not refer to spatial relationships. Library or image
catalogues are networks—forming edges between
the nodes of author and work—yet this is not a
spatial relation. In Schneider & Kohle’s analysis of
the “social image tagging” site Artigo (of which they
are both key contributors), they show how an
online catalogue of thousands of digital images may
be analyzed to reveal unforeseen clusters among
crowd-sourced tags. Using a clustering algorithm to
segment the tagged images into groups, they can
trace new categories of image proximity that may
expand traditional artistic genres (such as history,
portrait, still life, etc.). More than displaying visual
proximities, filtering images through crowdsourced annotations can generate new research
questions about classification itself and result in
new ways of framing images beyond their current
categorization. In the end, these subtler categories
may be used to train a Convolutional Neural
Network that could be made to automatically
classify digital images of artworks that are not
annotated or lack any metadata.
61The
2017); Art Institute of Chicago, Whistler and Roussel: Linked Visions,
http://linkedvisions.artic.edu/network.php (Accessed September 25, 2017).
62 Elijah Meeks, “More Networks in the Humanities or Did Books have DNA?” Digital
Humanities Specialist (blog), Stanford University, December 6,
2011: https://dhs.stanford.edu/visualization/more-networks/ .
Warhol, Warhol TimeWeb, https://vimeo.com/162073894 (Accessed
September 25, 2017); Smithsonian Archives of American Art, Six Degrees of Peggy
Bacon, http://www.smithsonianmag.com/smithsonian-institution/its-a-smallworld-after-all-six-degrees-of-peggy-bacon-806322/ (Accessed September 25,
ARTL@S BULLETIN, Vol. 6, Issue 3 (Fall 2017)
20
Visualizing Networks
Léa Saint-Raymond and Antoine Courtin similarly
engage questions about how digital art historians
can develop better methods for structuring their
data. Their research builds on studies of the art
market, which frequently apply quantitative
methods such as network analysis to better
understand the production, reception, and
exchange
of
art
among
interconnected
communities of producers and consumers. 63
Looking at the networks of artists that emerge from
archives of Parisian auction sales from the 19th and
early 20th centuries, Saint-Raymond & Courtin seek
to advance a better methodology for enriching their
data (to avoid flatness, imprecision, and overly
reductive data-sets), and also for cutting their data
(so that the network under analysis does not
become obscured by a proliferation of nodes and
edges). In order to enrich their data, they turned to
new and highly useful “linked open data platforms,”
in which networks become networked through the
semantic web. By enriching their data, their initial
corpus of auction records was expanded to create a
fuller picture of the social realities they trace.
However, after enriching, they also cut the data (by
removing “singletons” or artists that only appear
once in the record) in order to enhance their
articulation of the core network, which they
visualized using Gephi. Furthermore, as SaintRaymond & Courtin describe, at times they chose to
visualize the network not as a “network graph,” but
rather as a chart or table—both to better illustrate
the result of the analysis, and allow for a greater
level of interpretation.
structures that he employs to represent it. In
particular, their conversation explores how his
creative data visualizations disrupt the positivism
implied by contemporary infographics—exposing
how their claim to truth can obscure the
interpretive process embedded within them and
that gives them form. From statistical cartographies
of political affiliation and social sentiment, to
network diagrams of interpersonal communi-cation, DuBois explains how his works attempt to
interject indeterminacy, irony, and lyricism into
data analysis and visualization in order to
interrogate their assumed neutrality and
truthfulness. For example, Self-Portrait, 1993-2014
(2014)—a work DuBois made by running his email
through an algorithm and then cleaning, cutting,
and visualizing his data—was designed to render
an objective visualization of the artist’s social
network, and yet through the process, he found that
some of his more personal relationships were not
apparent due to the in-person or analogue nature of
much of their communications. Problems such as
data voids and reductive aggregates, as well as the
manner in which data visualization becomes
fetishized and gratuitously utilized, are explored in
the playful and provocative projects that DuBois
describes.
The second interview is a conversation that the
author conducted with Johanna Drucker and
Miriam Posner following the Getty/UCLA Summer
Institute that they co-organized, Beyond the
Digitized Slide Library.64 With the belated arrival of
the digital humanities to the discipline of art
history, the Summer Institute aimed to foster a
wider engagement with digital methods beyond the
digitizing of art works—teaching participants
about key debates, while offering hands-on training
with tools and techniques. Several of the
The final two contributions are interviews. In the
first of these two conversations, Anne Collins
Goodyear talks with the interdisciplinary and
intermedia artist, R. Luke DuBois, discussing how
data gains meaning with and through the graphical
Foundational studies in this field, include: Harrison and Cynthia White, Canvases
and Careers: Institutional Change in the French Painting World (New York: Wiley,
1965), and Robert Jensen, Marketing Modernism in Fin-de-Siècle Europe (Princeton,
NJ: Princeton University Press, 1997). Important recent articles include: Béatrice
Joyeux-Prunel, “Circulation and the Art Market,“ Journal for Art Market Studies Vol. 1,
no. 2 (2017): https://www.fokum-jams.org/index.php/jams/article/view/13/45
(Accessed September 25, 2017); Maximilian Schich, Christian Huemer, Piotr
Adamczyk, Lev Manovich, and Yang-Yu Liu, “Network Dimensions in the Getty
Provenance Index,” http://arxiv.org/abs/1706.02804 [physics.soc-ph] (Accessed
September 25, 2017).
64 This issue is indebted to the institute instructors Miriam Posner, Johanna Drucker,
Todd Presner, Steven Nelson, and Francesca Albrezzi, as well as all of the
participants and the fascinating projects that they brought to the workshop. The
content of the Summer Institute has been transformed into an online textbook and
collection of resources for those interested in gaining practical and theoretical
knowledge of Digital Art History. Posner, et al., Digital Art History 101 (Los Angeles,
CA: UCLA; Getty Foundation, 2016): https://ucla-beyond-slidelibrary.github.io/DAH101/About.html
63
Visualizing Networks
21
ARTL@S BULLETIN, Vol. 6, Issue 3 (Fall 2017)
contributors to this volume, the author included,
were participants in the institute, which
undoubtedly transformed each of our projects and
expanded our thinking about digital methods.
Referenced several times earlier in this
introduction, this interview serves as a kind of state
of the field of “digital art history.” Similar to the
discussions held during the Summer Institute, the
interview addresses key debates and points of
contestation, as well as exemplary projects within
the field. In the spirit of Beyond the Digitized Slide
Library, the aim of this issue is to further critical
dialogues about the visual tactics and theoretical
frameworks that structure quantitative and
computational approaches to art and its networks.
ARTL@S BULLETIN, Vol. 6, Issue 3 (Fall 2017)
22
Visualizing Networks